Standards
Editorial policy
How Fast Loading Themes selects subjects, separates evidence, makes recommendations, handles vendor input and corrects errors.
Last updated 16 August 2026
Our editorial promise
We write for the person choosing or maintaining a site, not for theme vendors or affiliate networks. Commercial relationships never purchase a ranking, evidence label, conclusion, recommendation or correction outcome.
Evidence categories
| Label | Meaning |
|---|---|
| Measured · lab | Collected by us under a published controlled protocol. |
| Field data | Aggregated real-user measurements, with source and scope identified. |
| Official fact | Version, update, requirements or features from a first-party source. |
| Developer claim | A vendor statement not independently established by our test. |
| User report | Anecdotal experience used as a research lead, not a benchmark fact. |
| Editorial analysis | Our interpretation, trade-off or recommendation. |
We do not silently promote one evidence class into another. In particular, repository-preview lab results are not presented as production performance or Core Web Vitals field data.
Selection and recommendation
Subjects are selected for user relevance, market use, distinct technical approaches, search demand, evidence availability and comparative value. A launch set may be intentionally focused rather than exhaustive. Inclusion does not imply endorsement; omission does not imply poor quality.
Measured rankings sort a named direct metric. Recommendation tools use separately explained fit logic around use case, editing model, commerce needs and dependency tolerance. We avoid unexplained composites that mix metrics, ratings, popularity and opinion.
Reviews, vendor access and free products
A vendor may provide technical clarification, product access or a review copy. That access will be disclosed when material. Vendors may identify factual errors before or after publication but do not approve wording or conclusions. If a vendor sponsors future research, sponsorship will be displayed prominently and the protocol, raw-data policy and editorial control will be stated before results.
Freshness and updates
Theme versions, prices, features and availability can change. Profiles carry source dates; measured data carries a collection date and environment. We target a 90-day review cycle for high-visibility benchmark evidence, but a displayed date—not an implied promise—defines currentness. Historical results should remain distinguishable from updated runs.
Corrections
We correct verifiable errors promptly. Material corrections that change interpretation, ranking or methodology receive a dated note in the affected page or methodology revision log. Minor spelling or formatting fixes may be made without a notice. Send the URL, disputed statement and primary evidence through our contact channels.